The effect of hypoxia was investigated in human (HPA) and rat (RPA) pulmonary arteries. Hypoxia-induced contraction was 95 +/- 8.7% and 9.3 +/- 4.8% of the control response to K(+)-rich (80 mM) solution in HPA and RPA, respectively (n = 10). When RPA strips were precontracted with phorbol 12,13 dibutyrate (0.2 microM), hypoxia elicited a larger contraction (105 +/- 13.4% of the control response, n = 8). In both types of artery, hypoxia-induced contraction was dependent on the extracellular calcium concentration (66 +/- 8.4% and 40 +/- 14.4%, reduction for 1.25 mM Ca2+ in HPA and RPA, respectively, n = 6) and was inhibited by verapamil (0.05-10 microM) and nifedipine (0.05-1 microM). Glibenclamide (5-10 microM) increased the amplitude of hypoxia-induced contraction (+42 +/- 5.3%, n = 5). Hypoxia-induced contraction was blocked by cromakalim (1 microM) and this effect was reversed by glibenclamide (5 microM). This contraction was also inhibited by iodoacetic acid (250 microM). In beta-escin skinned pulmonary arterial strips, hypoxia had no effect on the calcium concentration-tension relationship. These results suggest that the O2 sensor in the pulmonary artery is located on the vascular smooth muscle plasmalemma. Hypoxia-induced contraction is dependent on calcium influx through voltage sensitive calcium channels. Its amplitude is modulated by the functioning of potassium channels.